The Roundabout Society
The costs of complexity
In Austrian economics, the concept of capital is replaced by “roundabout production.” We can knead bread by hand, or we can produce kneaded bread in a roundabout way by obtaining a bread machine and using that to knead bread.
Modern economies use increasingly roundabout production. The processes for producing automobiles or smart phones involve countless steps and use parts that come from many places, including foreign countries.
As production becomes more roundabout, it becomes more complex. We become richer as a result of increasingly roundabout production. But the complexity changes our society in ways that are often unpleasant.
One consequence of complexity is a reduction in clarity and accountability. When a bad result comes from a simple process, the reasons for bad outcomes are transparent. It is easy to see what went wrong, to assign blame, and to prescribe fixes. When a bad result comes from a complex process, many steps could have been involved. Everyone can deny responsibility, and even an intense investigation may fail to determine the cause or suggest a solution.
Another consequence of complexity is a proliferation of formal rules. In a small business, you do not need an organization chart or an employee manual. In a large corporation, procedures must be written down and responsibilities specified as clearly as possible.
Every time a complex organization suffers a bad outcome, the response is to write down rules intended to prevent a recurrence. I like to say that an organization grows rules the way that an oyster grows pearls. Every pearl is a response to an irritation, trying to cover the source of the unpleasantness.
Formal rules are necessary. Without them, people work at cross-purposes and repeat the same mistakes. But rules also add to complexity and reduce flexibility.
When people complain that “we don’t build things any more” or “we are getting in out own way,” they make it sound as if complexity should not exist. When you first build a highway across farmland and prairie, it is relatively easy. After the land has become heavily settled, with many roads traversing the area, it becomes more difficult.
The proliferation of rules in a complex economy is like the proliferation of roads and buildings in a growing suburb. In hindsight, you can see how the rules are inefficient, but unraveling them is as difficult as re-engineering a road system that no longer can handle the increase in traffic.
As production gets more roundabout and the economy becomes more complex, the side effects of innovations and policy changes become more difficult to predict and control. Progress becomes more difficult.
For those of us who observe and kibbitz the regulatory apparatus, we would do well to recognize that the emergence of the roundabout economy is going to make solving problems more difficult. Saying that some things used to be easier 50 years ago is not really helpful. Many processes now work better. But complexity also has its costs.
Methodical Reform vs. Frenetic Hacking
A wise information systems manager once remarked to me that one law of computer systems is that “The person who created this system is an idiot.” Actually, the designers did a lot of things right, but later on new needs emerged and the system had to be patched. With enough patches, it starts to look as if it had been designed by an idiot. I was always a fan of frequently rewriting a system. I would rather pay the explicit cost of rewriting than suffer the subtle cost of “tech debt.”
But this systems manager failed. He scorned the idea of undertaking frenetic hacking But his methodical approach to rebuilding systems at our organization took too long to implement without yielding any rapid results.
One may think of Mr. Trump and Mr. Musk are dealing with the complex entity known as the U.S. government. If government were a computer system, it would consist of outdated code, with patches everywhere. It definitely behaves as if “The person who created this system is an idiot.”
One can think of two extreme ways of trying to tame such a monster. One way is to go about it methodically: gathering information, making a plan, and then executing the plan. The other way is to just start hacking away at it, working quickly and making a lot of mistakes in the process.
If you prefer the methodical approach, then it will be easy to criticize Mr. Trump and Mr. Musk. But keep in mind that the methodical approach also has its drawbacks. In particular, you may never get to the step of executing the plan. The current system has its defenders, which are a combination of people with a financial stake in the system and people who have as part of their identity a political loyalty to the Democratic Party and antipathy toward Republicans. While you are gathering information and making a plan, the system’s defenders will be creating obstructions, and your supporters will become demoralized by the lack of progress.
I wish that somehow we could get the best of both the methodical approach and the frenetic hacking approach. We need some quick wins to maintain morale, but given the roundabout society in which we live, we also need a methodical approach in order to effect meaningful change.


https://marginalrevolution.com/marginalrevolution/2025/03/peter-marks-forced-out-at-fda.html
It is proper to criticize DOGE as being the wrong way to improve the efficiency of the USG, but in reality one suspects that is not the objective at all, it's just acquiring power plus antipathy to many agency objectives. Musk is not really interested in making development aid more effective, improving the cost effectiveness of VOA, or reducing the overhead of sending out social security checks.
I appreciate the logic of this piece. It is true we need both methods. I just wish I saw any signs of laying long term plans for positive change. I can see the logic of trying to terrorize people in order to make your long term plans more palatable. Coincidentally (?) the next email in my inbox was from Marginal Revolution about Peter Marks (a key driver of Operation Warp Speed) resigning because of RFK’s measles craziness and the next one after that was about Canada’s opportunity to recruit the top US scientists coming out of university because our science funding is being slashed. Not what we were looking for I think.